Hi Steve,
Yes, indeed! I realised my mistake after I had sent my mail :(
I'm confused now because, I seem to have release 13.0.30 installed
and do have the missing file. I've chekced this and the variable
VO_ATLAS_SW_DIR on all my SL3 nodes. This is what I got on the node on
which a job failed:
[epcf37]# ls -l
/software/atlas/prod/releases/rel_13-2/AtlasOffline/13.0.30/AtlasOfflineRunTime/cmt/setup.sh
-rw-r--r-- 1 atlassgm atlas 610 Oct 16 13:47
/software/atlas/prod/releases/rel_13-2/AtlasOffline/13.0.30/AtlasOfflineRunTime/cmt/setup.sh
[epcf37]#
[epcf37]# env | grep VO_ATLAS_SW_DIR
VO_ATLAS_SW_DIR=/software/atlas
However, I do not have Release 13.0.30 on the SL4 nodes attached to my
second CE epgce2.ph.bham.ac.uk, but I'm not publishing it either and the
failures are with SL3 workers.
Do you know what could be the problem?
Thank you,
Yves
On Thu, 25 Oct 2007, Steve Lloyd wrote:
> Hi Yves,
> Do you mean my tests? The
> **** Warning: UI_CAN_NOT_EXECUTE ****
> Unable to execute "Python Tkinter Graphical":
> Unable to load library.
> comes from the UI and appears to be normal. The real error is a
> failure to access ATLAS release 13.0.30 although you are publishing that
> you have it:
> ./temp_UKI-SOUTHGRID-BHAM-HEP_MyHWPackage.sh: line 14:
> /AtlasOffline/13.0.30/AtlasOfflineRunTime/cmt/setup.sh: No such file or
> directory
> ./temp_UKI-SOUTHGRID-BHAM-HEP_MyHWPackage.sh: line 40: athena.py:
> command not found
> Cheers Steve
>
|